It is known that immune system, particularly T cells, plays an important role in tumor elimination by a living body. Indeed, infiltration of lymphocytes exhibiting cytotoxic effects on tumor cells in human tumor foci has been observed (Arch. Surg., 126:200, 1990), and cytotoxic T lymphocytes (CTLs) recognizing autologous tumor cells have been isolated from melanomas without great difficulties (e.g., Immunol. Today, 8:385, 1987; J. Immunol., 138:989, 1987; and Int. J. Cancer, 52:52, 1992). In addition, the results of clinical treatment of melanomas by transfer of the CTLs recognizing antologous tumor cells also suggest the importance of T cells in tumor elimination (J. Natl. Cancer. Inst., 86:1159, 1994).
Although it had long been unknown about target molecules for CTLs attacking autologous tumor cells, the recent advance in immunology and molecular biology gradually began elucidating such target molecules. Specifically, it has been found that CTL, using the T cell receptors (TCRs), recognizes a complex between a peptide, called tumor antigen peptide, and a major histocompatibility complex class I antigen (MHC class I antigen, and in the case of human, referred to as HLA antigen), and thereby attacks autologous tumor cells.
Tumor antigen peptides are generated by degradation of proteins specific for tumors, that is, tumor antigen proteins in cells with proteasomes, which proteins are intracellularly synthesized. The tumor antigen peptides thus generated bind to MHC class I antigens (HLA antigens) in endoplasmic reticulum to form complexes and the complexes are transported to the cell surface to be presented as an antigen. A tumor-specific CTL recognizes the complex presented as an antigen, and exhibits anti-tumor effects through its cytotoxic action or production of lymphokines. As a consequence of elucidation of a series of the actions, it has become possible to treat tumors by using tumor antigen proteins or tumor antigen peptides as so-called cancer vaccines to enhance tumor-specific CTLs in the body of a tumor patient.
As a tumor antigen protein, T. Boon et al. identified a protein named MAGE from human melanoma cells for the first time in 1991 (Science, 254:1643, 1991). Subsequently, several additional tumor antigen proteins have been identified mainly from melanoma cells. Examples of melanoma antigens that have been identified are melanosomal proteins such as a melanocytic tissue-specific protein, gp100 (J. Exp. Med., 179:1005, 1994), MART-1 (Proc. Natl. Acad. Sci. USA, 91:3515, 1994), and tyrosinase (J. Exp. Med., 178:489, 1993), MEGE-related proteins that are expressed not only on melanomas but also on various cancer cells and normal testicular cells (J. Exp. Med., 179:921, 1994), β-catenin having a tumor-specific amino acid mutation (J. Exp. Med., 183:1185, 1996), and CDK4 (Science, 269:1281, 1995). Tumor antigen proteins other than those from melanomas have also been identified, including products of oncogenes such as HER2-neu (J. Exp. Med., 181:2109, 1995) and p53 (Proc. Natl. Acad. Sci. USA, 93:14704, 1996), tumor markers such as CEA (J. Natl. Cancer Inst., 87:982, 1995) and PSA (J. Natl. Cancer Inst., 89:293, 1997), and viral proteins such as HPV (J. Immunol., 154:5934, 1995) and EBV (Int. Immunol., 7:653, 1995). Detailed descriptions of these subjects can be found in published reviews (e.g. Immunol. Today, 18:267, 1997; J. Exp. Med., 183:725, 1996; and Curr. Opin. Immunol., 8:628, 1996).
In applications of a tumor antigen protein or tumor antigen peptide to treatment or diagnosis of tumors, it is important to identify a tumor antigen that can be widely used for epithelial tumors such as gastric and lung cancers which occur at a much higher incidence compared to melanomas. In this relation, the present inventors conducted cloning of a gene encoding a novel tumor antigen protein from squamous carcinoma cells derived from esophageal cancer, and identified for the first time from tumor cells other than melanomas several tumor antigen peptides that are bound to and presented on HLA antigens of which HLA types are HLA-A24 or HLA-A26 (J. Exp. Med., 187:277, 1998; International Patent Publication WO 97/46676).
When these tumor antigen peptides are clinically applied in practice, it is desirable to use two or more different tumor antigen peptides rather than to use only a single peptide. That is to say, taking into consideration the facts that all cancer cells do not express an identical tumor antigen in common and that two or more different tumor antigen peptides are presented on a single cancer cell, a treatment using two or more different tumor antigen peptides is believed to be more effective. Indeed, in the case of melanoma, development of cocktail formulations comprising two or more peptides has been attempted, since a single peptide derived from a tumor antigen failed to exhibit adequate effects (Int. J. Cancer, 66:162, 1996; and Int. J. Cancer, 67:54, 1996). Under such circumstances, it is being required to identify novel tumor antigen proteins and tumor antigen peptides that can be widely used for epithelial tumors such as gastric and lung cancers which occur at a high incidence.